Performance Enhancement of a Coal Classifier

World energy demand is increasing relentlessly. The total global energy demand in 2030 is projected to be 50–60% above the current rate of energy consumption (IEA, 2008). Existing developed economies and fast-growing ones like China and India rely heavily on fossil fuels as a source of energy. Coal is still a key element in the energy mix for the world’s leading economies, and around 30% of all CO2 emissions come from the combustion of fossil fuels for electricity generation (IEA, 2008). Therefore, there is a need for clean coal technology to reduce the negative effect of the combustion. The coal particle size is critical to cleaner combustion; the classifier is responsible for that. The present work details an investigation into improving the performance of coal classifiers. The particular area of focus is to find the optimum design parameters by looking at the effects and influences of key classifier parameters towards the classifier performance. The use of ineffective classifier parameters, especially the vane angle and inlet velocity, reduces the performance of the classifiers where an inappropriate size of particle is being released. This contributes to a reduction in overall efficiency of the coal power plant and contributes to the formation of NOx gases during fuel burning. The performance of the classifier in terms of flow and particle distribution is the focus of the analysis. The work within this research study employs the Computational Fluid Dynamics (CFD) technique, which is a very effective, non-intrusive, virtual modelling technique with powerful visualisation capabilities. However, the importance of the experimental appreciative of the classifier is not neglected. Experiments were carried out to provide a tool for validating the CFD propositions. A one-third scale test facility that mimics an industrial air classifier has been carefully constructed in order to provide experimental data for the further understanding of the coal classification process. The outcome of this research work provides a guideline for selecting suitable parameters for specific classifier design and application.